 |
| Titel |
Changing the scale of hydrogeophysical aquifer heterogeneity characterization |
| VerfasserIn |
Daniel Paradis, Laurie Tremblay, Paolo Ruggeri, Patrick Brunet, Gabriel Fabien-Ouellet, Erwan Gloaguen, Klaus Holliger, James Irving, John Molson, René Lefebvre |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250103513
|
| Publikation (Nr.) |
 EGU/EGU2015-2928.pdf |
|
|
|
|
|
| Zusammenfassung |
| Contaminant remediation and management require the quantitative predictive capabilities of
groundwater flow and mass transport numerical models. Such models have to encompass
source zones and receptors, and thus typically cover several square kilometers. To
predict the path and fate of contaminant plumes, these models have to represent the
heterogeneous distribution of hydraulic conductivity (K). However, hydrogeophysics has
generally been used to image relatively restricted areas of the subsurface (small
fractions of km2), so there is a need for approaches defining heterogeneity at larger
scales and providing data to constrain conceptual and numerical models of aquifer
systems.
This communication describes a workflow defining aquifer heterogeneity that was applied
over a 12 km2 sub-watershed surrounding a decommissioned landfill emitting landfill
leachate. The aquifer is a shallow, 10 to 20 m thick, highly heterogeneous and anisotropic
assemblage of littoral sand and silt. Field work involved the acquisition of a broad range of
data: geological, hydraulic, geophysical, and geochemical. The emphasis was put on high
resolution and continuous hydrogeophysical data, the use of direct-push fully-screened wells
and the acquisition of targeted high-resolution hydraulic data covering the range of observed
aquifer materials. The main methods were: 1) surface geophysics (ground-penetrating radar
and electrical resistivity); 2) direct-push operations with a geotechnical drilling rig
(cone penetration tests with soil moisture resistivity CPT/SMR; full-screen well
installation); and 3) borehole operations, including high-resolution hydraulic tests and
geochemical sampling. New methods were developed to acquire high vertical resolution
hydraulic data in direct-push wells, including both vertical and horizontal K (Kv and
Kh).
Various data integration approaches were used to represent aquifer properties in 1D, 2D
and 3D. Using relevant vector machines (RVM), the mechanical and geophysical CPT/SMR
measurements were used to recognize hydrofacies (HF) and obtain high-resolution 1D
vertical profiles of hydraulic properties. Bayesian sequential simulation of the low-resolution
surface-based geoelectrical measurements as well as high-resolution direct-push
measurements of the electrical and hydraulic conductivities provided realistic estimates of the
spatial distribution of K on a 250-m-long 2D survey line. Following a similar approach, all
1D vertical profiles of K derived from CPT/SMR soundings were integrated with
available 2D geoelectrical profiles to obtain the 3D distribution of K over the study
area.
Numerical models were developed to understand flow and mass transport and assess how
indicators could constrain model results and their K distributions. A 2D vertical section
model was first developed based on a conceptual representation of heterogeneity which
showed a significant effect of layering on flow and transport. The model demonstrated that
solute and age tracers provide key model constraints. Additional 2D vertical section models
with synthetic representations of low and high K hydrofacies were also developed on the
basis of CPT/SMR soundings. These models showed that high-resolution profiles
of hydraulic head could help constrain the spatial distribution and continuity of
hydrofacies. History matching approaches are still required to simulate geostatistical
models of K using hydrogeophysical data, while considering their impact on flow
and transport with constraints provided by tracers of solutes and groundwater age. |
|
|
|
|
|
|